AI Center Training Problem

Feedback AI Center
,
1

Hey! Im trying to train the MultiLabelTextClassification model from UiPath Language Analysis but it keeps failling even when I use the JSON format from the UiPath Documentation that looks like this
[
{
“text”: “I like this movie”,
“label”: “7”
},
{
“text”: “I hated the acting”,
“label”: “9”
}
]
I deployed the Ml Skill, created the Pipeline and everytime it runs fails with this error.

Train only of rm_mini 23.5.0.0 launched - Run 24f4517d-405c-4608-951a-c43314d94f59
Train only of rm_mini 23.5.0.0 started - Run 24f4517d-405c-4608-951a-c43314d94f59
Train only of rm_mini 23.5.0.0 scheduled - Run 24f4517d-405c-4608-951a-c43314d94f59
Train only of rm_mini 23.5.0.0 failed - Run 24f4517d-405c-4608-951a-c43314d94f59

Error Details : Pipeline failed due to ML Package Issue

2024-01-19 13:17:36,550 - UiPath_core.trainer_run:main:83 - INFO: Starting training job…
2024-01-19 13:17:38,316 - UiPath_core.storage.azure_storage_client:download:118 - INFO: Dataset from bucket folder training-f462fcaa-d9d0-4819-b87e-8641a3ac6f15/f2b59cf1-00ee-4344-8c76-5895b23fe1da/2665faa9-4933-4a08-a1e1-24f771387021 with size 1 downloaded successfully
2024-01-19 13:17:38,316 - UiPath_core.training_plugin:train_model:130 - INFO: Start model training…
2024-01-19 13:17:38,316 - UiPath_core.training_plugin:initialize_model:124 - INFO: Start model initialization…
2024-01-19 13:17:38,317 - root:init:35 - INFO: Using options: {‘cuda’: True, ‘dataset’: {‘path’: ‘/microservice/dataset’, ‘csv_name’: None, ‘text_column_name’: ‘text’, ‘target_column_name’: ‘labels’, ‘delimiter’: ‘,’, ‘train_split’: 0.9, ‘encoding’: ‘utf-8’, ‘dataloader_workers’: 0}, ‘model’: {‘name’: ‘test’, ‘epochs’: 10, ‘tqdm’: True, ‘batch_size’: 24, ‘optimizer’: {‘transformer’: True}, ‘scheduler’: {‘type’: ‘hf’, ‘parameters’: {‘name’: ‘linear’}}, ‘gradient_accumulation_steps’: 1, ‘metrics’: [‘f1’, ‘ECC’], ‘path’: ‘/microservice/models/test’}, ‘network’: {‘name’: ‘roberta-base’, ‘freeze’: True}, ‘split_and_eval’: False}
2024-01-19 13:17:38,452 - UiPath_core.training_plugin:initialize_model:127 - INFO: Model initialized successfully
2024-01-19 13:17:38,452 - UiPath_core.training_plugin:model_run:179 - ERROR: Training failed for pipeline type: TRAIN_ONLY, error: No objects to concatenate
2024-01-19 13:17:38,453 - UiPath_core.trainer_run:main:100 - ERROR: Training Job failed, error: No objects to concatenate
Traceback (most recent call last):
File “/home/aicenter/.local/lib/python3.9/site-packages/UiPath_core/trainer_run.py”, line 95, in main
wrapper.run()
File “/microservice/training_wrapper.py”, line 58, in run
return self.training_plugin.model_run()
File “/home/aicenter/.local/lib/python3.9/site-packages/UiPath_core/training_plugin.py”, line 195, in model_run
raise ex
File “/home/aicenter/.local/lib/python3.9/site-packages/UiPath_core/training_plugin.py”, line 171, in model_run
self.run_train_only()
File “/home/aicenter/.local/lib/python3.9/site-packages/UiPath_core/training_plugin.py”, line 255, in run_train_only
self.train_model(self.local_dataset_directory)
File “/home/aicenter/.local/lib/python3.9/site-packages/UiPath_core/training_plugin.py”, line 132, in train_model
response = self.model.train(directory)
File “/microservice/train.py”, line 42, in train
self.process_data(split=True)
File “/microservice/train.py”, line 88, in process_data
df = pd.concat(dfs)
File “/home/aicenter/.local/lib/python3.9/site-packages/pandas/core/reshape/concat.py”, line 372, in concat
op = _Concatenator(
File “/home/aicenter/.local/lib/python3.9/site-packages/pandas/core/reshape/concat.py”, line 429, in init
raise ValueError(“No objects to concatenate”)
ValueError: No objects to concatenate
2024-01-19 13:17:38,453 - UiPath_core.trainer_run:main:107 - INFO: Job run stopped.

Does anyone has any feedback on this? why the “No objects to concatenate” error?

Thanks a lot

3

@Gomez_Heredero_Pablo

Looks like the second column is labels and not label…can you recheck

Cheers

1 Like
4

Hey Anil, it didnt work. I tried almost everything and is allways the same “No objects to concatenate” error. I have deployed the skills, run the pipeline and upload the dataset: a JSON file in a train folder.

What can i do?

5

@Gomez_Heredero_Pablo

If you have uploaded json select the type appropriately in the pipeline parameters

can you show dataset, sample json and the parameters in pipeline

cheers